Explore the Impact of the Application of Artificial Intelligence Facial

Image Processing and Synthesis Technology in Documentaries on

Social Acceptance

Bangjian You

Faculty Social Science and Humanities, The National University of Malaysia, Bangi, Selangor, Malaysia

Keywords: Artificial Intelligence, Documentary Film, Face Image Processing and Synthesis Technology, Technology

Ethics.

Abstract: In recent years, the widespread application of artificial intelligence technology in the field of face image

processing and synthesis has attracted widespread attention, and it has provided new technical means and

creative possibilities in the field of documentary production. However, due to the ethical risks of AI

technology application and the lack of social popularity, its application in documentaries is still controversial.

This paper analyzes the current status of the technical application of AI face image processing and synthesis

technology in the field of documentaries, the main problems it faces, and its impact on social acceptance. This

paper analyzes that although the application of AI face image processing and synthesis technology in the field

of documentaries has absolute practicality, the contradiction between its technical essence and "authenticity",

and the existing technical abuse problems existence will continue to affect its social acceptance in a short

period of time. Based on this, this paper puts forward the following suggestions: strengthen the popularization

and education of public AI knowledge, optimize the use of technology through government policies and

industry norms, and promote the improvement of AI technology detection and supervision system. Based on

the above solutions, the social acceptance of AI face image processing and synthesis technology and its

application effect in the field of documentaries can be improved to a certain extent.

1 INTRODUCTION

With the rapid development and widespread

application of artificial intelligence in various fields,

"face-changing" technology has provided new

thinking and aesthetic standards for the development

of the film and television industry with its precise

capture and realistic shaping, especially expanding

new areas of creation and improvement for

documentary media production. Artificial

intelligence is no longer just a background element of

documentaries. It is also closely related to the way

documentary makers create and distribute films

today. Artificial intelligence is quietly integrating into

all stages of documentary production - from research

and production to post-production and distribution

(Anlen & Cizek, 2024). However, as Craig Hight

pointed out, "documentary makers are considered

important stakeholders in the truth and trust non-

fictional forms of performance and participation."

The nature of AI face image processing and synthesis

technology obviously makes this technology

contradict the important principles of documentary

media. For many documentary creators,

documentaries are rooted in an idealism that desires

fragile responsibility, the truth of power, and the

authenticity of historical narratives (Cole, 2024).

Therefore, for some documentary producers and

directors, using this technology also faces significant

professional and social ethical issues. In response to

the irreversible impact of AI tools, the "Best Practice

Guidelines" of the Archives Producers Alliance

specifically emphasize the importance of protecting

primary sources and the need for transparency as AI-

generated images flood the media landscape, in order

to protect the integrity of archives, public records, and

collective memory (Anlen & Cizek, 2024). However,

the fact is that almost any documentary production

process cannot avoid the use of AI technology or

tools, because AI technology has been integrated into

almost all kinds of daily electronic products that

humans often come into contact with, including

mobile phones and cameras. The fact that AI has

become a common technology seems to be in serious

268

You, B.

Explore the Impact of the Application of Artiﬁcial Intelligence Facial Image Processing and Synthesis Technology in Documentaries on Social Acceptance.

DOI: 10.5220/0014112200004942

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on Applied Psychology and Marketing Management (APMM 2025), pages 268-272

ISBN: 978-989-758-791-7

conflict with the traditional concepts of the film and

television industry. At the same time, taking Fang Li's

film The Sinking of the Lisbon Maru (2024) as an

example, the content of the film contains a large

amount of research and production of historical

archival images, which is an indispensable part of the

production process for most documentaries,

especially historical documentaries. Many films that

refer to modern archives need to focus on the

acceptance of AI facial image processing and

synthesis technology by audience groups related to

historical figures in the film. This shows that there is

a certain degree of conflict between the practicality

and ease of use of this technology in historical film

and television works and the acceptance and ethical

standards of the audience in real situations. As Floridi

and Luciano mentioned, the ethical risks of AI face-

changing technology, including privacy violations

and moral disputes, may significantly reduce the

audience's acceptance of it (Davis, 1989). In the face

of different views and opinions put forward by

creators and audiences, it is absolutely necessary to

explore the relationship between the use of AI face

image processing and synthesis technology and social

acceptance. Therefore, based on the above-stated

questions, this paper will analyze: Is AI face image

processing and synthesis technology suitable for use

in documentary production, what types of problems

will AI face image processing and synthesis

technology encounter during use, and what methods

can solve the current problems of AI face image

processing and synthesis technology?

Based on these three main research questions, this

paper will analyze the characteristics and

connotations of documentary production based on

existing literature and film and television work cases,

explore the significant and potential problems of AI

face image processing and synthesis technology in the

current documentary application process, and give

corresponding solutions. At the same time, this paper

mainly divides social acceptance into documentary

creator acceptance and audience acceptance. By

analyzing the two separately and answering the above

research questions, it explores the impact of the

application of this technology in documentary media

on social acceptance.

2 CASE STUDY

After the introduction and improvement of artificial

intelligence tools in recent years, the rapid

development of AI has changed the relevant fields of

Hollywood science fiction movies and all aspects of

the video game industry, especially in the simulation

of human 3D avatars. Today, the frequency of

application of artificial intelligence in the field of

documentaries is not reduced compared with science

fiction films and games because of the pursuit of real

records in documentaries. On the contrary, the

usefulness and ease of use of AI face image

processing and synthesis technology in the restoration

of documentary archive images have made it very

popular in the process of documentary production

(Karras et al., 2019). So, in an era where AI

technology has become a common technology,

whether AI technology should be used and how to use

it correctly are issues that documentary producers

need to explore.

According to the articles on AI face image

processing and synthesis technology in the two

searches of EURASIP Journal on Image and Video

Processing and MDPI, the current research focuses

on: face image restoration and completion, face

replacement (deep fake technology), facial

expression recognition and synthesis, deep fake

detection and defense, multi-dimensional and multi-

modal face synthesis, etc. Among them, the

technologies widely used in documentary production

include face repair and completion technology, face

swapping technology, deep fake detection

technology, and 3D face reconstruction technology.

Most of the technologies mentioned above basically

serve the restoration and enhancement of

documentary archive images. Especially for

documentaries on historical themes, the application of

this technology has opened up a new creative field for

them and greatly improved the creation ceiling of

documentaries on this theme. They Shall Not Grow

Old directed by Peter Jackson is a successful case of

the use of AI face image repair and completion

technology in documentaries. AI technology is used

on a large scale to repair specific character images

and reconstruct 3D faces for old and damaged

historical archive images, which undoubtedly greatly

improves the completion of the film and the

audience's viewing experience. Other movies that use

the same technology as this work include One Day

When We Were Young (2021). In addition, the movie

Here (2024) directed by Robert Zemeckis is a model

of the use of face swapping technology.

Metaphysical.ai specifically used this technology to

age and de-age the real-time starring Tom Hanks in

the movie. The StyleGAN architecture was first

proposed by Timo Alia's research team in 2018 and

presented in a paper in 2019 (Lees, 2023). This

architecture is a combination of multimodal synthesis

technology and 3D face construction technology. It

Explore the Impact of the Application of Artiﬁcial Intelligence Facial Image Processing and Synthesis Technology in Documentaries on

Social Acceptance

269

can generate realistic face images in documentaries,

animations, and games, and can even construct

completely virtual faces.

The above cases show that AI face image

processing and synthesis technology has developed

rapidly in recent years, and may bring more practical

technical models and architectures to the

documentary industry in the future. At the same time,

these cases also once again confirm the importance of

AI technology in the documentary production

process. The current application of AI technology has

provided documentary producers with more space for

artistic creation. Andy Warhol Diaries (2022) narrates

this four-hour series through the deep sound of

Warhol "reading" his diary. Such a unique processing

method reflects the artistic quality of the work

pursued by some documentary producers. Compared

with traditional documentary documentaries, some

creators try to focus on showing the "specialness" of

their works to the audience through more ingenious

processing methods. This pursuit of artistry even

includes the transition of different film styles or the

interweaving of visual types. For example, in addition

to interviews with people, historical impact

broadcasts, and historical data displays, One Day

When We Were Young also restores the state of mind

of historical figures in disguised form through

animation, giving the boring historical documentary

a unique poetic feel while addressing real-life issues

from a specific perspective, giving the work a

stronger ideological depth, criticality, and

educational inspiration. It can be seen that as an

artistic collection of multiple characteristics and

connotations, the production restrictions of

documentaries continue to expand with the upgrading

of AI technology. To a certain extent, AI technology

can make it easier for documentaries to achieve

qualified production goals in the minds of creators.

3 PROBLEM ANALYSIS

Based on the literature and case analysis presented

above, it can be proved that AI face image processing

and synthesis technology has a great promoting effect

on documentary production. However, this

technology still has some urgent problems to be

solved in the current creative environment, especially

the acceptance of this technology by different social

roles. For some documentary producers who have

achieved successful applications, AI face image

processing and synthesis technology is a tool with

significant practicality and ease of use, but for

producers who hold the opposite opinion, the

application of AI technology, especially the

application of a large area, is not allowed. What's

more, for some audience groups who have a shallow

understanding of this technology and the

documentary production process, they are prone to

relatively serious psychological barriers to the use of

some AI technologies. Therefore, analyzing the

problems existing in the current application of AI face

image processing and synthesis technology in

documentaries and its social impact will help to

change the society's acceptance of this technology in

the future.

3.1 Social Morality and Professional

Ethics in Documentary Film

As mentioned above, some documentary creators try

to pursue the historical authenticity that

documentaries originally have. The core reason why

they oppose the use of artificial intelligence

technology is that AI challenges the perception of

reality, as well as the construction, narration and trust

of truth (Anlen & Cizek, 2024). Although such

incidents are not new problems for the film and

television industry, as Yuval Noah Harari mentioned,

the speed and scale of media output by artificial

intelligence technology are exponential, and the lack

of human supervision is unprecedented. Therefore,

this group of creators believes that AI has caused a

new information crisis, and the facial image

processing and synthesis technology in AI technology

has even more immeasurable destructive power, and

the harm caused by this technology has caused

obvious fluctuations in human society. The

prevalence of deep fake technology is a concrete

manifestation of the use of AI facial image processing

and synthesis technology in opposition to the basic

ethical and moral rules of human society. For

documentary creators, being disapproved and

condemned by the audience or even the public for

using this technology in their works is not a

consequence they want to face. The professional

ethics of documentaries often attach great importance

to the moral responsibility and honesty of the creators

to the audience (Wang et al., 2024). In addition, as

mentioned above, the current human society lacks

absolute supervision over the development and use of

AI technology. Behind all kinds of social harm caused

by deep fakes is the neglect of personal privacy

issues. All technologies related to human biometric

recognition or construction, including AI face image

processing and synthesis technology, always have

issues involving personal privacy. For example,

Anthony Bourdain was condemned for narrating

APMM 2025 - International Conference on Applied Psychology and Marketing Management

270

short texts in his private emails without revealing it

through a deep fake voice generated by artificial

intelligence in his film Roadrunner (2021). Therefore,

some social groups will have a certain degree of

resistance to the application of AI technology in

documentaries.

3.2 Audience Acceptance

Based on the technology acceptance model, perceived

usefulness and perceived ease of use are the two main

determinants of the technology acceptance model,

and these determinants are positively correlated with

the user's acceptance attitude towards a certain

technology (Venkatesh & Davis, 2000). Perceived

usefulness refers to the extent to which users believe

that using a certain technology or system can improve

their work performance or life efficiency, while

perceived ease of use refers to how little effort users

believe is required to learn and use a certain

technology or system. However, since most

documentary viewers have little understanding of the

core of AI face image processing and synthesis

technology and the documentary production process,

the perceived usefulness and perceived ease of use of

AI technology cannot be intuitively experienced by

the audience. In addition to perceived usefulness and

perceived ease of use, user usage attitudes may be

affected by external factors, including ethical

considerations and risk perception, which in turn

affect adoption decisions (Dinev & Hart, 2006).

Therefore, when they cannot intuitively feel the

advantages of AI technology in documentaries, most

viewers will focus more on referring to their own

ethical risk perceptions. The Perceived Ethical Risk

and Perceived Privacy Concern scales proposed by

Dinev & Hart roughly subdivide ethical risk

perceptions (Floridi & Taddeo, 2016). Its content

includes but is not limited to the following

dimensions: unknown risks, moral risks, privacy

risks, authenticity risks, social impact risks, etc. In

addition to the audience's concerns about their own

privacy and the audience's pursuit of the authenticity

of the film mentioned above, the audience's

acceptance is directly affected by the audience's

limited understanding of the technology, which will

lead to active avoidance or only follow the self-

ethical risk cognition standards. Therefore, for

documentary audiences, the real viewing experience

of the film is proportional to the audience's

acceptance of AI face image processing and synthesis

technology, and the audience's acceptance of AI face

image processing and synthesis technology is

proportional to their understanding and recognition of

the technology.

3.3 Technical Limitations

Taking the face-changing technology in AI facial

image processing and synthesis technology as an

example, the technology first appeared in 2014, and

the corresponding software appeared in 2018 and was

officially put on the market. With the appearance of

OpenAI in the past two years and its extensive use,

AI biometric technology has gradually entered the

public eye. In 2023, the application of AI facial image

processing and synthesis technology ushered in a

staged climax. Although the development speed of

artificial intelligence has a steady upward trend in

recent years, and the upper limit of the research and

development of this technology is immeasurable.

However, with only 6 years of research and

development history, the technical maturity of AI

biometric technology is still questionable. At the

same time, the complexity of the use of AI facial

image processing and synthesis technology has a very

differentiated upper and lower limits. For users who

have not learned professional knowledge, they do not

have the ability to fully master it. Therefore, the

current ease of use of this technology has not reached

the expected level.

4 SUGGESTION

The problems faced by AI face image processing and

synthesis technology in the current documentary

creation environment mainly stem from the fact that

humans have a relatively shallow overall

understanding and control of AI technology. As a

product of the new media era, the characteristics and

development trends of artificial intelligence are an

impact on traditional media. The short time of

research and development and contact has directly

affected some of the technologies and concepts of

traditional media and failed to match AI technology,

which has indirectly led to the low tolerance of human

society for this technology. Although the acceptance

of new technologies by the times and society will

increase year by year with the advancement of time,

and new technologies and old media will gradually

merge, the promotion of more technical knowledge

and the limitation of the scope of technology use to

avoid more "abuse of deep fake technology" are

indispensable actions in the process of forming this

result. In the process of solving the problem of

imperfect understanding of AI technology in human

Explore the Impact of the Application of Artiﬁcial Intelligence Facial Image Processing and Synthesis Technology in Documentaries on

Social Acceptance

271

society, popularizing basic knowledge and improving

people's AI literacy are considered the primary

method (Scantamburlo et al., 2024).

At the same time, several ethical issues embodied

by AI technology in the current social environment

will gradually be solved with the development and

improvement of technology. AI facial image

processing and synthesis technology is currently

developing exponentially, and many existing

technical and ethical issues are being constantly

resolved and replaced. The implementation of AI

deep fake detection technology has to some extent

made up for the technical defects and loopholes of AI

face-changing technology in the early stage. In

documentary production, it can realize multi-program

detection and repair, thereby basically ensuring the

stiffness and distortion of face-changing technology

in documentary applications. The two films Welcome

to Chechnya (2023) and Another Body (2023) use

artificial intelligence technology to protect the

identities of vulnerable subjects by changing (rather

than blurring) their faces and voices. For the

audience, this makes the subject humane, protects

their well-being, and fully interprets the two-sided

nature of deep fake technology. Therefore, when

facing the social and ethical issues of AI technology,

using AI deep fake detection technology to control

the use of AI technology and controlling the use of AI

facial image processing and synthesis technology

through systematic management and control within

the basic ethical and moral scope of human society is

a way to solve this problem (Seth, 2024).

5 CONCLUSION

In the analysis of this study, it can be concluded that

the impact of the application of artificial intelligence

face image processing and synthesis technology in

documentaries on social acceptance is objectively

changing with the maturity of technology, social

opinion and time. A large number of successful cases

have proved that the application of AI face image

processing and synthesis technology in documentary

production is absolutely practical, and it is very

suitable for the production of documentaries and

other types of film and television works without

touching the bottom line of ethics. The application of

AI face image processing and synthesis technology in

documentaries starts with a small and specific focus,

and indirectly presents the overall acceptance and

focus of artificial intelligence technology in human

society. With the iteration and update of technology,

the AI deep fake technology that has deeply disturbed

human society may only become a symbol of an era

or a reference case for future technological

development in the future.

REFERENCES

S. Anlen, K. Cizek, The synthesis: AI is everywhere, even

in your documentaries. Int. Doc. Assoc. (2024).

W. Cole, The dilemma of documentary proof: Archival

producers alliance co-founders discuss how AI will

change archival. Int. Doc. Assoc. (2024).

F. D. Davis, Perceived usefulness, perceived ease of use,

and user acceptance of information technology. MIS Q.

13(3), 319-340 (1989).

T. Karras, S. Laine, T. Aila, A style-based generator

architecture for generative adversarial networks. In

Proceedings of the IEEE/CVF Conference on

Computer Vision and Pattern Recognition 4401-4410

(2019).

D. Lees, Deepfakes in documentary film production:

Images of deception in the representation of the real.

Inf. Fusion (2023).

A. Wang, Y. Zhou, H. Ma, X. Tang, S. Li, R. Pei, M. Piao,

Preparing for aging: Understanding middle-aged user

acceptance of AI chatbots through the technology

acceptance model. Digital Health. 10, 1-10 (2024).

V. Venkatesh, F. D. Davis, A theoretical extension of the

technology acceptance model: Four longitudinal field

studies. Manag. Sci. 46(2), 186-204 (2000).

T. Dinev, P. Hart, An extended privacy calculus model for

e-commerce transactions. Inf. Syst. Res. 17(1), 61-80

(2006).

L. Floridi, M. Taddeo, What is data ethics? Philosophical

Transactions of the Royal Society A: Mathematical,

Physical and Engineering Sciences 374(2083),

20160360 (2016).

T. Scantamburlo, A. Cortés, F. Foffano, C. Barrué, V.

Distefano, L. Pham, A. Fabris, Artificial intelligence

across europe: A study on awareness, attitude and trust.

IEEE Trans. Artif. Intell. (2024).

J. Seth, Public perception of AI: Sentiment and opportunity

arXiv preprint arXiv:2407,15998 (2024).

APMM 2025 - International Conference on Applied Psychology and Marketing Management

272